One or more aspects relate, in general, to processing within a computing environment, and in particular, to improving such processing.
In computing, floating point numbers are used to represent real numbers, and there are a number of representations used for floating point numbers. These representations include indications of how certain operations are to behave with floating point numbers. For example, the Institute of Electrical and Electronics Engineers (IEEE) provides a floating point standard which specifies the behavior for minimum and maximum operations on floating point numbers.
Although this standard exists, many programming languages and programmers have chosen not to follow this standard. Thus, there are several different standards and representations among the various programming languages within their libraries. Even within a language, the common convention may be to not use a library function, but to use some other language syntax that leads to differing behavior. This leads to difficulty in accelerating these operations in hardware and often requires a sequence of many instructions including several conditional branches to emulate.